原发性肝癌预后因素的Cox模型分析研究  

NALYSIS OF PROGNOSTIC FACTORS OF PRIMARY LIVER CANCER BY COX REGRESSION MODELS

作　　者：黄天壬 张振权 罗斯[2] 莫志纯[2] 

机构地区：[1]广西肿瘤研究所 [2]广西医学院

出　　处：《广西医科大学学报》1993年第1期26-29,共4页Journal of Guangxi Medical University

摘　　要：应用Cox模型对523例原发性肝癌患者的预后进行分析研究,结果从31个可能影响的因素中,筛选出临床分期、手术切除、放疗、肿瘤大小、治后AFP改变、临床分型、插管化疗、肿瘤结节数目等8个对肝癌预后有影响的因素。据此建立肝癌生存数学模型及预测模型,并对每例患者计算预测值,估计其预后。预测值估计预后与实际预后相比较,灵敏度、特异性及符合率分别为83.8％、92.7％和90.8％。523 patients with PLC were analysed with the Cox Regression Models to determine the main prognostic factors of Primary Liver Cancer(PLC).31 variables were used in the analytic procedure and 8 were found to be of significance.These eight factors included the clinical stage,suigical resection,radiotherapy,tumor size,changes of the serum alpha-fetoprotein,clinical type,infusion chemotherapy,and the number of tumor nodules.A prognostic index was constructed.The predicting values(PV)of the patients were calculated according to the above patients' individual features.The predicting values and the actual values were compared,and the sensitivity,specificity and overall concordance achieved were 83.8%,92.7% and 90.8% respectively.

关 键 词：原发性肝癌 预后 COX回归模型 

分 类 号：R[医药卫生]